Thread: Gehry Technologies Digital Project

The main difference between 3D modeling tools and BIM tools is the fact that BIM tools recognize objects as opposed to recognizing geometry (mathematical description of planes (nurbs) or an array of XYZ points (polygons)).


Let’s take a simple form such as a cube as an example.


In a pure 3D modeler, to create a cube, one would specify 8 points in XYZ space. The planes which are derived from these points (which are only visual representations- the software does not really “see” the planes unless they are nurbs) will create the visual representation of the cube. Furthermore the cube in this example has a zero depth (with either double sided or single sided normals), which means that it has no “substance”. By contrast this cube – if it was to be modeled in BIM toll like ArchiCAD for instance – would be made out of two slab objects (where each object is a kind of a flat cube in itself) and four walls (which again are actually tall cubes)


Now, even though the appearance of both cubes is the same, from the point of view of geometry, the BIM model contains 6 x 8 = 48 XYZ points VS the 8 of the 3D modeler. Not to mention the fact that the BIM tool actually created – what would be from the stand point of the 3D modeler – a cube within a cube!
This of course makes for a much more cumbersome model, but this is for a good reason! The data in the BIM model is “intelligent”. In other words; the software knows that a slab is a slab and a wall is a wall and can therefore assign certain attributes for these components, which correspond to their attributes in real life; Which is the main point behind BIM!


Therefore the main power of BIM is also it shortcoming, i.e. inability to construct complex shapes (especially compound curves, such as Ghery’s designs)


There is a solution of course. Model the complex shapes in a pure 3D modeler and import them into a BIM application for limited analysis. Why limited? Because even though the BIM application will be able to recognize the geometry, it would not be able to associate this geometry with a particular intelligent object (such as Walls for instance) and instead will treat it as a generic object (such as GDL in the case of ArchiCAD)
This will entail additional work during the construction documentation phase mainly due to the need for extra 2D work on the appearance of the GDL object in the various views (such as sections for instance)


I’m not sure to what extent other useful data such as volume or other quantities can be derived out of these generic objects, but I’m sure some basic coding can do the work if one has the basic programming knowledge.


For conclusion, there is no doubt that the combination of; a pure 3D modeler with a BIM tool, is extremely powerful; and worth the effort of devising a good link between them as well as developing a smooth geometry pipeline.